Microbial Energizers: Fuel Cells That Keep on Going Microbes that produce electricity by oxidizing organic compounds in biomass may someday power useful electronic devices

Derek R. Lovley

as this happened to you? You have ganisms with the ability to oxidize organic com- a layover between flights, would pounds to while transferring like to use your computer and cell electrons to electrodes with extraordinarily high H phone, but both sets of batteries efficiencies. Electricigens make it possible to are drained and the nearby electri- convert renewable biomass and organic wastes cal outlets are being used. What if you could directly into electricity without combusting the instead recharge your electronic devices with a fuel, which wastes substantial amounts of en- little sugar from the nearby coffee stand? With ergy as heat. Efforts to eliminate the inefficien- help from electricity-producing microorgan- cies of combustion are behind the recent interest isms, known as electricigens, some day you in hydrogen fuel cells, which oxidize hydrogen might have new options for ignoring the current and reduce oxygen to water while producing “grid” by generating electricity in an alternative, electricity in a controlled chemical reaction. environmentally friendly manner. With electricigens, however, it becomes pos- Electricigens are recently discovered microor- sible to make microbial fuel cells, which offer potential advantages over hydrogen fuel cells. For example, hydrogen fuel cells re- quire a very pure source of a highly explo- Summary sive gas that is difficult to store and distrib- • Electricigenic microorganisms such as ute. Furthermore, hydrogen is derived Geobacter and Rhodoferax efficiently oxidize mainly from fossil fuel rather than renew- organic compounds to carbon dioxide while able sources. In contrast, the energy sources directly transferring electrons to electrodes. for microbial fuel cells are renewable organ- • Electricigen-based microbial fuel cells mark a ics, including some that are dirt cheap. paradigm shift because these cells completely oxidize organic fuels while directly transferring electrons to electrodes without mediators. Geobacteraceae Producing • Although microbial fuel cells are unlikely to produce enough electricity to contribute to the Electricity in Mud national power grid in the short-term, the cells Several years ago, Leonard Tender of the may prove feasible in some specific instances such as covering the local energy needs for pro- Naval Research Laboratories in Washing- Derek R. Lovley is cessing food wastes. ton, D.C., and Clare Reimers of Oregon Distinguished Uni- • Optimizing microbial fuel cells will entail devel- State University in Corvallis developed sys- versity Professor oping a better understanding of how electron tems in which electricigens produce electric- and Director of En- transfers occur along the outer surfaces of elec- ity from mud! When a slab of (the vironmental Bio- tricigens; key challenges include increasing an- anode) is buried in anaerobic marine sedi- technology at the ode surface areas and increasing electricigen ments and then connected to another piece University of Mas- respiration rates. of graphite (the cathode) that is suspended sachusetts, Am- in the overlying aerobic water, electricity herst.

Volume 1, Number 7, 2006 / Microbe Y 323 F I G U R E 1

Sediment fuel cell. (A) Prior to deployment in salt marsh sediments on Nantucket Island, Mass. (B) Diagram of sediment fuel cell reactions. (C) Deployed sediment fuel cells. (Photos courtesy of Kelly Nevin, University of Massachusetts-Amherst.)

flows between them (Fig. 1). Although this ar- Which Geobacteraceae prove to be prevalent rangement typically produces meager electrical in such samples depends on the specific environ- currents, they are adequate for running analytic ment being tested. For example, if electrodes are monitoring devices similar to those that investi- placed in marine sediments, Desulfuromonas gators place in remote locations such as the predominate, whereas if the electrodes ocean bottom. are placed in freshwater sediments, Geobacter How do such sediment fuel cells produce elec- species predominate. Although Geobacter and tricity? The simple answer is, with microbes. Desulfuromonas species have similar physiolo- Dawn Holmes, working with Daniel Bond in gies, Desulfuromonas prefer marine salinity, my laboratory, scraped the anode surface with while Geobacter favor freshwater. a razor blade, extracted DNA from those A hallmark of Geobacteraceae is their ability scrapings, and determined what species were to transfer electrons onto extracellular electron present based on their 16S rRNA genes. The acceptors. For example, Geobacter and Desul- surprising result is that such anodes are highly furomonas species support growth by coupling enriched with microorganisms in the family the oxidation of organic compounds to the re- Geobacteraceae. When similar pieces of graph- duction of Fe(III) or Mn(IV) oxides. Further- ite are incubated in sediments but not connected more, these microorganisms can transfer elec- to a cathode in overlying water, there is no such trons to other metals and to the quinone enrichment. moieties of humic substances, which are so large

324 Y Microbe / Volume 1, Number 7, 2006 that they must be reduced outside bac- F I G U R E 2 terial cells. Reducing Fe(III) oxides is an important means for degrading organic matter in aquatic sediments, submerged soils, and subsurface environments. Mo- Calculator –! –! lecular analyses of such environments re- 8 e 8 e veal that, in general, Geobacteraceae are the predominant Fe(III)-reducing micro- organisms in zones in which Fe(III) re- e–! duction is important. 8 Holmes and Bond found that Outlet 2 CO2! 2 O2! O in Geobacteraceae can also use electrodes +! +! 2 + 8 H! + 8 H! as extracellular electron acceptors. Geobacter 8 e–! Both Desulfuromonas and Geobacter –! +! 8 e 8 H species can grow by oxidizing organic ! compounds to carbon dioxide, with Fuel in Outlet electrodes serving as the sole electron Acetate! 4 H2O (C H O )! acceptor. Moreover, more than 95% of 2 4 2 + 2 H2O the electrons derived from oxidizing such organic matter can be recovered as electricity. In sediment fuel cells, Geobacteraceae oxidize organic com- Anode chamber Cathode chamber pounds but, instead of transferring elec- Cation-selective! trons to Fe(III) or Mn(IV), their natural membrane electron acceptors, they transfer elec- Schematic of Geobacter-powered . trons onto electrodes (Fig. 1). The elec- trons flow through the electrical circuit to the cathode, where they react with oxygen to form water. would go to the electricigenic microbe via aero- bic respiration. However, the electricigens still recover some energy from electron transfer to Self-Perpetuating, Highly Efficient, the electrode. This energy recovery is very im- Geobacter-Based Microbial Fuel Cells portant because the energy that the electricigens The sediment fuel cell can be recreated with pure conserve allows them to maintain viability and cultures of Geobacter (Fig. 2). The anaerobic to produce electricity as long as fuel is provided. anode chamber contains organic fuel and a Nearly a century ago, M. C. Potter at the graphite electrode. The cathode chamber has a University of Durham in England measured similar electrode and is aerobic. Geobacter electrical currents when electrodes were placed transfers electrons released from oxidized or- in microbial cultures. In this and other studies ganic matter onto the anode. The electrons flow carried out throughout much of the 20th cen- from the anode to the cathode. The two cham- tury, microbes generated electricity by produc- bers are separated by a cation-selective mem- ing soluble, reduced compounds that could react brane that permits the protons that are released abiotically with electrode surfaces. In initial from oxidized organic matter to migrate to the studies these were natural reduced end products cathode side, where they combine with electrons of fermentation or such as and oxygen to form water. hydrogen, sulfide, alcohols, or ammonia. How- The cation-selective membrane limits oxygen ever, many of these reduced products react only diffusion to the anode chamber, preventing slowly with electrodes, and other end products, Geobacter from oxidizing the organic fuels with such as organic acids, do not appreciably react the direct reduction of oxygen. By inserting an with electrodes at all. Adding soluble electron electrical circuit within the flow of electrons to acceptors, known as electron shuttles or media- oxygen, energy can be harvested that otherwise tors, enhances current production in such sys-

Volume 1, Number 7, 2006 / Microbe Y 325 F I G U R E 3 Geobacter-Based Fuel Cells Mark a Paradigm Shift Although a few years ago fuel cell ex- perts thought that direct electrochemi- cal contact between microorganisms and electrodes was virtually impossible, this mechanism appears to be how Geo- bacteraceae carry out electron transfer to electrodes. Thus, the use of Geo- bacteraceae in microbially based fuel cells marks a paridigm shift. They com- pletely oxidize organic fuels to carbon dioxide while directly transferring elec- trons to electrodes without mediators. There has been no known evolutionary pressure on microorganisms to produce electricity. Therefore, it is hypothesized that Geobacter cells transfer electrons to electrodes via the same mechanisms that they use when reducing extracellu- lar, insoluble electron acceptors, such as Fe(III) oxides, that they encounter in Transmission electron micrograph of Geobacter covering graphite anode. (Photo cour- natural environments. tesy of Daniel Bond, University of Massachusetts-Amherst.) Evidence for direct electron transfer from Geobacteraceae to electrodes comes from a variety of studies. For instance, Kelly Nevin at UMASS-Amherst dem- onstrated that G. metallireducens has to directly tems. These electron shuttles enter cells in the contact Fe(III) oxides to reduce them. Daniel oxidized form, accept electrons from respiratory Bond found that the cells of closely related G. components within the cell, exit in reduced form, sulfurreducens that attach to electrode surfaces and donate electrons to an electrode, which (Fig. 3), rather than planktonic cells, are respon- recycles them into the oxidized form. However, sible for producing power in microbial fuel cells. there are drawbacks to using such mediators— Electrochemically active proteins on the outer they add expense to electricity production, and surface of G. sulfurreducens could serve as elec- many of them are toxic to humans and/or unsta- trical contact points between the microbes and ble. Mediators are especially unsuitable for elec- electrode surfaces. tricity-generating strategies in open environ- If the electrode is adjusted to a low enough ments. Furthermore, the microbes used in these potential, it can act as an electron donor for systems typically were fermentative and thus Geobacter species, rather than an electron accep- most of the electrons available in the organic tor, according to Kelvin Gregory in my lab. Labo- fuel remained in organic products instead of ratory studies have suggested that this process being transferred to the electrodes. might be used to provide Geobacter with electrons More recently, studies in the laboratory of to remove contaminants, such as , from Byung Hong Kim at the Korea Institute of Sci- polluted water via reductive precipitation. ence and Technology demonstrated that fuel cells containing species could pro- Electricigens Other than Geobacteraceae duce electricity from lactate without the addi- Microorganisms outside the Geobacteraceae tion of electron shuttles. However, the efficiency family can also oxidize organic compounds to of electron transfer was low in part because carbon dioxide, with electrodes serving as the Shewanella species only incompletely oxidize sole electron acceptor. For example, Swades lactate to acetate. Chaudhuri from my lab found that Rhodoferax

326 Y Microbe / Volume 1, Number 7, 2006 ferrireducens can completely oxidize sugars capable of reducing Fe(III), it may be possible to with electron transfer to electrodes. produce electricity under extreme conditions. Sugars are important constituents of many Most notably, the capacity for reducing Fe(III) is wastes and renewable biomass. Although highly conserved among hyperthermophilic bac- Geobacter species oxidize a variety of organic teria and archaea. acids and aromatic compounds as well as hydro- gen, none appears to oxidize sugars. Therefore, Potential Practical Applications producing electricity from sugars with for Fuel Cells Geobacter species also requires fermentative mi- croorganisms to convert those sugars to organic The primary near-term practical application of acids and hydrogen. Rhodoferax offers the pos- fuel cells powered by electricigens is likely to be sibility of directly converting these sugars to sediment fuel cells designed to power electronic electricity with a single organism. monitoring equipment in remote locations. In sediment fuel cells that we tested in fresh- However, electricigens can extract electricity water sediments, we detected 16S rRNA gene from a wide range of other sources of microbi- sequences on the anodes that appear closely ally degradable organic wastes or renewable related to Geothrix fermentans, although at biomass. Although oxidizing these organic fuels much lower levels than Geobacter sequences. G. yields carbon dioxide, this process returns only ferementans is an acetate-oxidizing Fe(III) re- recently fixed carbon to the atmosphere and ducer, and Daniel Bond found that G. fermen- thus is not a net contributor to atmospheric tans can also oxidize acetate with the produc- carbon levels. Furthermore, oxidizing these ma- tion of electricity. The G. fermentans cells terials in fuel cells would produce none of the appear to be enmeshed in an extracellular ma- pollutants usually associated with combustion. trix on the electrode, in contrast with When wastes are the energy source, potential Geobacter-covered electrodes, which carry lit- environmental contaminants are consumed tle, if any, extracellular material. We speculate while producing electricity. that Geothrix produces this material to limit Kelvin Gregory in my lab showed that micro- losses of an electron shuttling compound it re- bial fuel cells can convert swine wastes to elec- leases and that the high energetic cost of produc- tricity, avoiding the usual waste-handling pro- ing a shuttle limits the ability of Geothrix to cess that releases methane and odor-causing compete with Geobacter species on electrodes. organic acids. In his studies, Geobacteraceae In marine sediments with high concentrations accounted for more than 70% of the microbes of sulfide, electrodes may also be colonized by living on the surface of anodes that were im- microorganisms in the family Desulfobul- mersed in the swine waste. baceae, according to my colleague Dawn Meanwhile, Willy Verstraete at Ghent Uni- Holmes. Sulfide can react directly with elec- versity in Ghent, Belgium, and Bruce Logan at trodes, where it is oxidized to elemental sulfur. Pennsylvania State University in University Desulfobulbus propionicus, a Fe(III)-reducing Park, among others, are designing reactors for representative of the Desulfobulbaceae, can ox- efficiently converting high volumes of animal idize elemental sulfur to sulfate with an elec- wastes and human sewage into electricity. trode as the electron acceptor. Thus, when sul- Which microorganisms are producing electricity fate reducers are actively involved in degrading in these systems is not well understood, but organic matter in marine sediments, sulfide organisms other than Geobacteraceae typically might serve as an electron carrier that can be predominate. generated at a distance from the electrode sur- Microbial fuel cells that produce enough elec- face—providing electrons for electricity from tricity from organic wastes are unlikely to sub- both abiotic and biotic reactions. stantially contribute to the national power grid It seems likely that many other types of micro- in the short term. Not only would such a system organisms can directly transfer electrons to elec- be an engineering marvel but, even if optimized, trodes, and some of them may have properties it would be difficult to compete with other with practical significance. Furthermore, if the sources of relatively cheap electricity, such as capacity for direct electron transfer to electrodes fossil fuels and nuclear fission. Nonetheless, mi- is a general characteristic of microorganisms crobial fuel cells may prove practical sooner for

Volume 1, Number 7, 2006 / Microbe Y 327 F I G U R E 4 Fuel Cells Need Optimizing before Applications Become Common Why are some of these applications not yet in place? For one thing, electricigens were discovered only recently. For an- other, they produce power slowly, suit- able for low-energy devices such as simple calculators (Fig. 4) or as trickle- charging devices for traditional batter- ies (see www.geobacter.org). In order for microbial fuel cells to power a wider assortment of electronic devices, the cells will need to oxidize fuels more rapidly than they now can. A key design challenge is to increase anode surface areas because of the di- rect relationship between anode surface area and power output. Other electro- chemical considerations include ensur- ing that internal resistances and oxygen reduction rates at the cathode do not restrict electron flow. Geobacter fuel cells powering a calculator. (Photo courtesy of Kelly Nevin, University of Optimizing microbial fuel cells will Massachusetts-Amherst.) also entail developing a better under- standing of how electricgens transfer electrons from their outer surface onto anodes. As we learn more about the electrical contacts between microbes and electrodes, we some relatively high-energy liquid wastes, such can begin to develop materials for electrodes as those from processing food or milk, where that better interact with the electron transfer electricity generation could help to cover treat- proteins of the electricigens. Moreover, we ment costs. can perhaps genetically engineer these microbes Another short-term practical application to produce more or better contacts with elec- could be the powering of electronic devices trodes. without connecting them to the grid—espe- We are evaluating several outer-membrane cially, say, in developing countries where micro- proteins that might serve as electrical contact organisms are widely used to convert domestic points between G. sulfurreducens and fuel-cell waste to methane gas that is used locally for electrodes. One candidate is a highly abundant cooking. Converting such wastes to electricity c-type cytochrome, OmcS, that is displayed on instead of methane would provide greater versa- the outside of the cell. Teena Mehta in my lab tility. Another possibility is to develop aquatic demonstrated that OmcS is required for extra- or terrestrial “gastrobots,” robots that consume cellular electron transfer onto Fe(III) oxides. organic matter to power their locomotion and Another candidate is pili, according to Gemma sensing and computational needs. Reguerra in my lab and Kevin McCarthy and Meanwhile, Bruce Rittman at Arizona State Mark Tuominen in the University of Massachu- University and his collaborators are evaluating setts-Amherst Physics Department. They dem- whether microbial fuel cells can be designed to onstrated that G. sulfurreducens pili are electri- use astronaut wastes as an electric energy source cally conductive and function as microbial during space travel. More down to earth, other nanowires (Fig. 5). Genetic studies and the phys- engineers are considering whether microbial ical location of the pili suggest that they can fuel cells could provide energy for mobile elec- serve as the final conduit for electron transfer tronic devices or automobiles. between the cell and the Fe(III) oxides.

328 Y Microbe / Volume 1, Number 7, 2006 Another path to increasing the elec- F I G U R E 5 tricity output of microbial fuel cells may be to increase Geobacter’s respiration rate. Mounir Izallalen from my laboratory and Radhakrishnan Mahadevan at Genomatica, Inc., in San Diego, Calif., used a genome-based model of G. sulfurreducens to formu- late a strategy for increasing its respira- tion rate. They then used genetic engi- neering to produce cells that respired faster. These efforts to understand how Geobacter and other electricigens pro- duce electricity come when market forces encourage development of smaller, more efficient electronic de- vices as well as alternative sources for increasingly costly fossil fuels. Hence, further study of electricigens not only will provide valuable insights into the elegance of extracellular electron trans- Transmission electron micrograph of the abundant electrically conductive pili of fer but could also lead to novel engi- . (Photo courtesy of Gemma Reguera, University of Massa- neering concepts that bring practical chusetts-Amherst.) benefits to consumers.

SUGGESTED READING Bond, D. R., D. E. Holmes, L. M. Tender, and D. R. Lovley. 2002. Electrode-reducing microorganisms harvesting energy from marine sediments. Science 295:483–485. Chaudhuri, S. K., and D. R. Lovley. 2003. Electricity from direct oxidation of glucose in mediator-less microbial fuel cells. Nature Biotechnol. 21:1229–1232. Gregory, K. B., and D. R. Lovley. 2005. Remediation and recovery of uranium from contaminated subsurface environments with electrodes. Environ. Sci. Technol. 39:8943–8947. Logan, B. E. 2005. Simultaneous wastewater treatment and biological electricity generation. Water Sci. Technol. 52:31–37. Lovley, D. R. 2006. Bug juice: harvesting energy with microorganisms. Nature Rev. Microbiol., in press. Rabaey, K., and W. Verstraete. 2005. Microbial fuel cells: novel biotechnology for energy generation. Trends. Biotechnol. 6:291–298. Reguera, G., K. D. McCarthy, T. Mehta, J. Nicoll, M. T. Tuominen, and D. R. Lovley. 2005. Extracellular electron transfer via microbial nanowires. Nature 435:1098–1101. Shukla, A. K., P. Suresh, S. Berchmans, and A. Rahjendran. 2004. Biological fuel cells and their applications. Curr. Sci. 87:455–468.

Volume 1, Number 7, 2006 / Microbe Y 329